Skip to content

Latest commit

 

History

History
123 lines (84 loc) · 4.98 KB

README.md

File metadata and controls

123 lines (84 loc) · 4.98 KB

Cloudflare Worker for GrowthBook Integration

This Cloudflare Worker integrates with GrowthBook to dynamically handle requests, track experiments, and manage user attributes. The worker also appends the fromWorker=true query parameter to requests and manages cookies for tracking anonymous users and cached events.

Features

  • GrowthBook Integration: Manages user attributes and tracks experiment views using GrowthBook.
  • Cookie Management: Sets cookies for tracking user sessions, cached events, and other attributes.
  • Query Parameter: Appends fromWorker=true to all outgoing requests to indicate Worker origin.
  • Geo-location Handling: Extracts country information from the request using Cloudflare's cf.country.
  • Flexible Deployment: Configured for both staging and production environments.

Environment Variables

The Worker uses the following environment variables:

Variable Description
GROWTHBOOK_CLIENT_KEY The GrowthBook client key for making API requests.
GROWTHBOOK_DECRYPTION_KEY (Optional) The decryption key for GrowthBook data.
GROWTHBOOK_API_HOST The API host for GrowthBook (e.g., https://api.growthbook.io).

How It Works

  1. The Worker checks if the fromWorker query parameter is present. If not, it appends fromWorker=true to the URL.
  2. It extracts cookies such as rudder_anonymous_id, cached_analytics_events, and utm_data from the request headers.
  3. It tracks user interactions and experiments with GrowthBook.
  4. It sets cookies for session tracking and stores experiment results.
  5. It manages user attributes such as device type, language, country, and UTM data.

Setting Up

1. Install Dependencies

Ensure you have installed the required dependencies for running the worker (e.g., GrowthBook, uuid, cookie parser).

If using npm:

npm install

2. Configure Wrangler

Set up your wrangler.toml configuration file to handle both staging and production environments.

Example wrangler.toml:

name = "my-worker"
type = "javascript"
account_id = "your-cloudflare-account-id"
workers_dev = true
compatibility_date = "2024-10-03"

# Staging environment
[env.staging]
vars = { ENVIRONMENT = "", PROXY_TARGET="", GROWTHBOOK_CLIENT_KEY = "", GROWTHBOOK_API_HOST="" }
route = "https://staging.example.com/*"

# Production environment
[env.production]
vars = { ENVIRONMENT = "", PROXY_TARGET="", GROWTHBOOK_CLIENT_KEY = "", GROWTHBOOK_API_HOST="" }
route = "https://example.com/*"

3. Set Environment Variables

Use Cloudflare's environment settings to configure your environment variables for both staging and production:

# Staging environment
wrangler secret put GROWTHBOOK_CLIENT_KEY --env staging
wrangler secret put GROWTHBOOK_DECRYPTION_KEY --env staging
wrangler secret put GROWTHBOOK_API_HOST --env staging

# Production environment
wrangler secret put GROWTHBOOK_CLIENT_KEY --env production
wrangler secret put GROWTHBOOK_DECRYPTION_KEY --env production
wrangler secret put GROWTHBOOK_API_HOST --env production

4. Deploy to Staging and Production

To deploy to the staging environment:

wrangler publish --env staging

To deploy to the production environment:

wrangler publish --env production

5. Testing

You can test the worker by making a request to your domain with and without the fromWorker=true query parameter to verify that the Worker is bypassing the redirect rules when necessary.

Cookie Management

The following cookies are managed by the Worker:

Cookie Name Description
rudder_anonymous_id Tracks the anonymous user session.
cached_analytics_events Stores cached events related to experiment tracking.
utm_data Stores UTM parameters for campaign tracking.
client_information Stores client information to verify if the user is logged in.

Query Parameters

The Worker appends the fromWorker=true query parameter to the request URL to signal that the request is coming from the Worker and to bypass any global redirect rules that might interfere with the Worker’s execution.

Geo-Location

The Worker uses Cloudflare’s request.cf.country to extract the client's country information and uses this information for tracking purposes and to customize the user experience.


Example Usage

Visit https://deriv.com to trigger the worker. If you want to bypass the redirect rule for testing purposes, visit https://deriv.com/?fromWorker=true.